Local antibiotics have a role in orthopedic trauma for both infection prophylaxis and treatment. They provide the advantage of high local antibiotic concentration without excessive systemic levels. Nonabsorbable polymethylmethacrylate (PMMA) is a popular antibiotic carrier, but absorbable options including bone graft, bone graft substitutes, and polymers have gained acceptance. Simple aqueous antibiotic solutions continue to be investigated and appear to be clinically effective. For established infections, such as osteomyelitis, a combination of surgical debridement with local and systemic antibiotics seems to represent the most effective treatment at this time. Further investigation of more effective local antibiotic utilization is ongoing.
Local antibiotics have the advantage of high local concentrations (thus efficacy at the surgical site), and low systemic concentrations (less risk of systemic side effects).
Local antibiotics have been proven effective for infection prophylaxis and treatment of established infection, and are typically used in concert with systemic antibiotics.
Multiple delivery systems are available for antibiotic delivery, with each having unique properties that may be advantageous.
Antibiotic delivery from PMMA is highly variable and depends upon: surface area (bead size), antibiotic used, number of antibiotics, mixing technique, time since implantation, Fluid characteristics around the beads, and others.
Aqueous antibiotic solution injected locally after wound closure is a simple delivery method that has demonstrated positive results in animal and clinical models.
Local antibiotic use began more than 100 years ago with Joseph Lister, who pioneered safe, antiseptic surgery. Before Lister’s innovations, as many as 80% of all operations were complicated by infection. He was the first to apply local antiseptics, including carbolic acid, to surgical wounds to treat open fractures. This led to further use of local antiseptics by Fleming during World War I, and in 1939 with Jensen instilling sulfanilamide crystals as local antibiotic in open fractures for infection prevention. Despite significant advances in the use of prophylactic antibiotics and perioperative protocols, orthopedic surgical site infections still remain a significant source of morbidity and mortality and result in a substantial financial burden to the health care system. Surgical site infections are the second most common cause of nosocomial infections in extra-abdominal surgeries, with an incidence of 2% to 5%, and approximately 5% of orthopedic internal fixation implants becoming infected. The rate of infection following internal fixation of closed fractures is generally much lower than that of open fractures, with open fractures approaching 30%. Despite the higher infection rate seen in the treatment of particular fractures compared with arthroplasty, there is much less literature available on the prophylactic use of local antibiotics for infection prevention in open and closed fracture treatment.
Local antibiotics provide high local concentrations with lower systemic levels than parenterally administered antibiotics. The delivery of local antibiotics can both supplement and sometimes obviate the need for systemic antibiotics. In certain instances, the target area of treatment may be avascular, preventing systemic antibiotics from reaching the targeted site. In these scenarios, local antibiotics may serve as the only effective option in treating the infection. Perhaps the main advantage of local antibiotic therapy is the ability of an antibiotic to reach a high local concentration while simultaneously having a low or undetectable systemic concentration, thereby avoiding certain negative side effects, such as nephrotoxicity and ototoxicity and decreasing the chances of developing pathogenic resistance. At this high level of local concentration, many bacteria that might otherwise be normally resistant to an antibiotic fall within its spectrum of activity.
In addition to infection prophylaxis, local antibiotics may have a role for treatment of established infections. This antibiotic therapy is typically coupled with surgical debridement when necessary, which includes wide excision of infected and devascularized tissues, curettage of abscesses and sequestra, restoration of soft tissue coverage, and removal of all foreign bodies. Although these techniques help to eradicate infection, they also contribute to the formation of dead space. Various antibiotic carriers can help fill and manage this potential space caused by bone or soft tissue defects, preventing subsequent development of infection ( Table 1 ).
|Recommended Mixture||Trade Examples||Other Considerations|
|PMMA||2 g vancomycin and 2.4 g tobramycin in 40 g PMMA cement||Palacos (Zimmer, Warsaw, IN), Simplex (Stryker, Kalamazoo, MI), SmartSet (DePuy, West Chester, PA)||May take longer to set up and may require additional monomer when additional antibiotics are added|
|Calcium sulfate||1 g vancomycin and 1.2 g tobramycin in 10 mL packet of calcium sulfate||Osteoset (Wright Medical, Memphis, TN), Stimulan (Biocomposites, Wilmington, NC)||FDA approved as a bone void filler, antibiotic delivery is off-label use; adding tobramycin powder only after mixing CaSO 4 will help it set up|
|Aqueous solution||80 mg tobramycin in 40 mL solution||Available as generic tobramycin, prepared in the OR||Inject into wound AFTER wound closure; if a drain is in place, clamp drain while injecting solution|
One potential negative implication of a high local concentration of antibiotics is cytotoxicity, which could inhibit new bone formation and delay fracture union at high enough levels. We will review commonly used carriers and methods for local antibiotic administration, their indications, and recent clinical trials evaluating the success of these methods.
One aspect of treating infection involves isolating the pathogen from the infected tissue or bone and determining the sensitivity of that pathogen to different antimicrobial agents. This goal is most readily accomplished when treating the unicellular, planktonic bacteria that are present in an infected wound bed. Conversely, biofilms interfere with this strategy. A biofilm is an extracellular matrix produced by bacteria that offers protection and provides an organizing scaffold to facilitate metabolic activity and communication between the bacteria within the matrix ( Fig. 1 ). In a biofilm, bacteria may tolerate antibiotic concentrations up to 1000-fold greater than the same bacteria in planktonic form. Biofilm bacteria are not as mobile or virulent within the body as their unicellular phenotypes; however, they are much more protected from host immunity and systemic antibiotics and thus more difficult to eradicate.
Once established, the biofilm can provide a continual source of bacteria that can detach as planktonic cells or biofilm fragments that can then travel to and infect other sites or cause a systemic infection. Even though they are less virulent, biofilms cause damage by invoking a host inflammatory response that generates adjacent tissue destruction, manifesting clinically as pain and implant loosening. Biofilms, and the bacteria that comprise them, have the ability to attach to orthopedic implants through their unique surface structures. The most common biofilm-producing organisms found in orthopedic infections are Staphylococcus aureus, coagulase-negative staphylococci, beta-hemolytic streptococci, and aerobic gram-negative rods, such as Pseudomonas aeruginosa .
Unfortunately, failure to isolate an organism from an implant-related infection is not uncommon. Although the planktonic bacteria can be isolated and grown with traditional culture techniques, identifying bacteria from biofilms is often unsuccessful. Sonication of explanted implants has been shown to improve the sensitivity of intraoperative cultures. Even with accurate intraoperative cultures, established infections often require formal irrigation and debridement for infection eradication or suppression, depending on the acuity of the infection. In these situations, the ability to eradicate the infection depends on removal of biofilm at the infection site, often through extensive resection and formation of potential space.
Recent research has focused on the prevention of biofilm formation via modification of implants to alter the surfaces, inhibiting bacteria adhesion. One example using this strategy has been the coating of voice prostheses with silicone rubber. Although this has been successful in the field of otolaryngology, these materials may interfere with osseointegration of orthopedic implants, limiting its applicability. Other strategies have focused on covalently linking antibiotics to the implant surface. Translational research has shown this technique to be effective in inhibiting S aureus implant colonization while still supporting bone healing in a large animal model. However, there is concern that this novel technique may encourage and even stimulate bacteria to develop resistance. Regardless of the implant, all medical devices are susceptible to biofilm colonization and infection. Strategies for the treatment and prevention of biofilm formation must be considered to reduce the morbidity and cost of orthopedic implant infections.
The ideal local antibiotic delivery system has yet to be formulated, but would produce a high local antibiotic level at the target site and concurrently allow a safe systemic level. The elution rates, factors that influence elution rates, and the interaction between the environment and the material would need to be defined. The material would be easily handled and manipulated, removed if nonabsorbable, and nonimmunogenic and inexpensive. For an absorbable material, it would need break down in a relatively short time, such that it did not act as a foreign body once the antibiotic was eluted. In the past 2 decades several different local antibiotic delivery carriers have been used. These can largely be divided into 2 groups based on the biodegradability of the delivery vehicle.
Antibiotic-loaded bone cement may be considered the current gold standard for local antibiotic delivery in orthopedic surgery. Antibiotic-loaded polymethylmethacrylate (PMMA) cement beads are the most popular nonbiodegradable modality used in conjunction with surgical debridement and systemic antibiotic therapy and have been used to treat and prevent bone and soft tissue infections for almost 30 years.
Antibiotic-loaded PMMA can be applied in multiple settings for the treatment and prophylaxis of infection. Common indications include the prevention of infection in total joint arthroplasty, open fractures, and the management of potential space (dead space) in patients with large bone or soft tissue deficits. It also can be used to treat acute and chronic osteomyelitis, chronic infected nonunions, and periprosthetic joint infections.
Contraindications are largely limited to patient hypersensitivity or allergy to specific antibiotics as well as the presence of resistant organisms such as Enterococcus . The presence of the beads themselves is an attractive surface for slime-producing organisms, such as Enterococcus , and this slime barrier decreases the efficacy of the antibiotic. The theoretic advantages of antibiotic beads include a high local concentration with low systemic levels, occupation of potential space following surgical debridement, low immunogenic response, and a high surface area of the bead allowing for a rapid release of the antibiotic.
The surgical technique involves mixing the antibiotic powder with the powdered cement polymer and then adding the methylmethacrylate liquid monomer. The cement is then inserted into a bead mold or beads can be formed by hand. They are then typically connected with either 26-gauge wire or nonabsorbable heavy suture. The author’s preferred technique is to use a 0 Prolene suture and pass the suture through the beads as they are hardening, thus making the beads easier to place and later retrieve at the time of removal.
Regarding the mechanical effect of antibiotic on the PMMA, biomechanical testing performed on Smart Set GHV (DePuy Orthopedics Inc, Warsaw, IN) and CMW 1 (CMW Laboratories Ltd, Devon, United Kingdom) to determine if their structural properties was compromised with the addition of antibiotics (linezolid, gentamycin, vancomycin, linezolid plus vancomycin, linezolid plus gentamicin). With up to 2 g antibiotic per 40 g PMMA packet (5% weight/weight), there was no reduction in the axial compression strength of each brand of cement. However, 4.5 g or greater of antibiotic powder has been shown to weaken PMMA. One method to maintain or increase the mechanical strength of the cement is to vacuum-mix the batch, which will reduce porosity and thus increase strength.
The antibiotic used must be water soluble, available in powder form, be chemically stable, and have a broad antibacterial spectrum with a low percentage of resistant species. The antibiotic must also be thermally stable, as the polymerization of the cement is an exothermic reaction creating temperatures up to 60 to 80°C. The most commonly mixed antibiotics that fit the above profile are gentamicin, tobramycin, and vancomycin. The aminoglycosides are effective against aerobic gram-negative bacilli and staphylococci in addition to streptococci, enterococci, and anaerobes. Tobramycin is more commonly used in the United States due to its wide availability as a pharmaceutical-grade power. Vancomycin can be added when the risk of resistant staphylococcal organisms is present. Vancomycin has been shown to be heat resistant and is readily available in powder form, with effective elution properties.
Although elution is ultimately governed by the difference in the concentration of antibiotic in the cement and its surrounding environment, other factors affecting elution include the type and viscosity of PMMA, the type and concentration of the antibiotic, and the structural characteristics of the beads. Increasing the surface area–to-volume ratio (ie, smaller beads) increases the elution of antibiotics. The type of antibiotic also affects elution, with tobramycin able to elute antibiotics longer and sustain concentrations above the minimum inhibitory concentration for longer periods of time than vancomycin at the same dose. Moreover, elution of antibiotics from PMMA beads has been extensively studied and remains a debated issue that is not completely understood. In general, there is a biphasic release pattern that occurs with an initial rapid release of approximately 5% to 7% of the total amount of antibiotic released within the first 24 hours, followed by a sustained secondary elution of antibiotic that steadily decreases over weeks or months. In fact, elution has been reported up to 5 years after PMMA bead implantation. Ultimately, multiple factors contribute to the elution profile of an antibiotic from PMMA beads, making it quite difficult to standardize the system for consistent antibiotic delivery.
Antibiotic-loaded PMMA beads also can be administered in an antibiotic bead pouch ( Fig. 2 ). With this technique, antibiotic-loaded beads are placed into a bony or soft tissue defect, and the wound is not closed, but is covered with an occlusive dressing, such as Ioban (3M, St. Paul, MN) ( Figs. 3 and 4 ). Negative-pressure wound therapy (NPWT) may be used in conjunction, although this is at the surgeon’s discretion and has produced somewhat conflicting results in recent studies. Although Stinner and colleagues showed decreased efficacy of bead pouch used with NPWT, Warner and colleagues found that for extremity blast injuries, compared with a Vacuum-Assisted Closure Therapy system (KCI Inc, San Antonio, TX), a bead pouch resulted in less late methicillin-resistant Staphylococcus aureus infections, although more unanticipated returns to the operating room for wound problems, and the bead pouch group required more surgeries overall until closure of the wounds.
Several animal and clinical studies have been conducted that have supported the beneficial role of PMMA beads in the prevention of infection following bone contamination. Fitzgerald and colleagues demonstrated a 90% prevention rate in the development of osteomyelitis following contamination with S aureus after insertion of gentamicin-loaded cement, whereas Chen and colleagues reported a significant reduction in the bacterial count of S aureus after the insertion of tobramycin-loaded beads in a rabbit model. In one of the largest clinical trials, Ostermann and colleagues compared the addition of an antibiotic bead pouch versus systemic antibiotics alone in preventing infection in 1085 open fractures. The group reported infection rates of 3.7% in those treated with the antibiotic bead pouch in addition to systemic antibiotics, compared with 12% in those treated with systemic antibiotics alone. Furthermore, several animal and clinical studies have investigated the potential therapeutic applications of PMMA beads in the treatment of osteomyelitis. A randomized controlled study by Calhoun and colleagues suggested that long-term systemic antibiotic therapy following debridement and reconstructive surgery for infected nonunions can be substituted by local antibiotic therapy in the form of PMMA beads. Patzakis and colleagues further corroborated the effectiveness of PMMA beads after demonstrating a 100% union rate in patients with chronic osteomyelitis and bony defects who underwent debridement, systemic antibiotics, and bead placement.
Several different doses of antibiotics in PMMA have been reported in the orthopedic literature. Most arthroplasty studies report a range of 1 to 2 g of vancomycin and 2.4 to 3.6 g of tobramycin. Although there is no consensus, the author’s preferred dosage is 2 g vancomycin and 2.4 g tobramycin in 40 g PMMA cement.
Intramedullary nails made from antibiotic-loaded PMMA cement may be used for antibiotic delivery in long-bone infections. Two approaches are commonly used, each having its advantage. A nail composed of PMMA can be fabricated with a 40-French chest tube, which has an 11-mm inner diameter. A 16-gauge or 18-gauge Luque wire is commonly placed in the middle of the nail to allow removal if it breaks. The advantage of this technique is maximal antibiotic delivery, as there is more cement present, and it is often used when stability is not the primary objective. The other technique involves coating a titanium nail (either pediatric nail or standard 9-mm tibia nail) with antibiotic-loaded PMMA cement. This may be either formed by hand, or injected into a chest tube around the nail. If done by hand, the interlock screw holes may be preserved. This technique allows some antibiotic to be delivered to the target site, but its primary advantage is stability. For either technique, removal of the chest tube surrounding the antibiotic nail is facilitated by cooling the nail in a water bath and coating the inside of the chest tube with mineral oil before filling with cement.
One additional application of antibiotic-loaded PMMA cement is the Masquelet technique, which is a 2-stage strategy for the reconstruction of segmental diaphyseal defects. Developed in 1986, the technique takes advantage of induced membranes to reconstruct the defects with nonvascularized bone autograft. The first stage includes standard debridement, and insertion of PMMA antibiotic-loaded cement spacer into the defect, and closure or coverage of soft tissue. The second stage occurs 6 to 8 weeks later after definitive healing of the soft tissue envelope has occurred. The spacer is taken out with careful attention paid to not disrupt the membrane that has been induced by the cement. This cavity now surrounded by the membrane is packed with cancellous bone autograft that can be combined with demineralized bone matrix to fill the void. The technique relies on the theory that the biological membrane induced by the PMMA cement has a protective and positive effect on the cancellous autograft. A recent retrospective study reported on 84 posttraumatic diaphyseal long-bone reconstructions using the technique over a 20-year period. The series was composed of largely open fractures (89%) of which union was obtained in 90% at a mean of 14 months after the first stage of the reconstruction. The investigators report that the technique provides a successful way to manage segmental defects and control infection before bone reconstruction.
Controversies concerning PMMA beads and other forms of nonbiodegradable local antibiotic therapy include length of implantation and the need for removal. Prolonged implantation may lead to the development of drug-resistant bacteria. Despite killing glycocalyx-forming bacteria during the elution phase, bacteria can persist and adhere to the retained PMMA beads, now acting as foreign bodies, and may survive on their surface after release of antibiotic has fallen below therapeutic levels. This adherence might provide an environment for recurrence and resistance, as has been seen in the wounds of patients that have been treated with gentamicin-loaded acrylic cement beads in the past. One center has reported an increase in the prevalence of resistant bacteria with the introduction of antibiotic-loaded bone cement. PMMA itself has been associated with decreased immune function and response, which might further impair the eradication of any remaining infection. In response to these concerns over prolonged implantation, bead removal within 4 to 6 weeks from implantation has been recommended because the beads progressively become incorporated within callus and entrapped in fibrous tissue, which likely reduces elution and can complicate retrieval.
In summary, PMMA beads have proven to be an effective nonbiodegradable option for local antibiotic therapy for prevention and treatment of infections in open fractures and osteomyelitis. For maximal effectiveness, the author recommends that the beads be used as an adjunct in the care of infection and not as a substitute for debridement.
Although PMMA may be considered the gold standard for local antibiotic delivery for the prevention and treatment of orthopedic infection, concerns over the use of PMMA has led to the investigation of alternate biodegradable materials as delivery vehicles. Concerns include variable release properties of PMMA and retained PMMA acting as a foreign body after antibiotic release falls below therapeutic levels, creating a surface for reinfection and bacterial resistance. In contrast, a biodegradable implant with a faster, complete, release of antibiotic would theoretically decrease the risk of recurrence of infection and generation of resistance. Furthermore, biodegradable implants obviate the need for a second surgery for removal. An osteoconductive, bioabsorabable bone substitute that is clinically as effective as PMMA in infection eradication has several clinical advantages. Biodegradable antibiotic delivery vehicles can be broadly grouped into 4 different categories: bone graft, bone graft substitutes or extenders, natural polymers, and synthetic polymers.
Bone autograft and allograft, combined with antibiotics, have been used clinically for more than 2 decades as a delivery vehicle to treat infection. The concept was developed in 1984 and centered on using a material that was already required for the reconstruction at the time of a secondary surgery to remove the antibiotic-laden cement. The morselized bone incorporates during bone regeneration and remodeling, allowing the recruitment of host defenses to protect the now vascularized bone graft and zone of previous infection. Antibiotics are added as a powder to the morselized bone autograft or allograft or the graft can be soaked in an antibiotic solution. The antibiotic is absorbed directly to the bone surface, and release is known to occur through first-order kinetics. In vitro and in vivo studies in a rabbit model demonstrated first-order kinetics for release of tobramycin and vancomycin over a period of 3 weeks with levels exceeding usual bactericidal concentrations. In a clinical study, tobramycin and vancomycin levels were studied in 26 patients with antibiotic morselized cancellous bone grafts greater than 20 mL. Data demonstrated continued release for at least 3 weeks, safe serum levels, and drain fluid levels 10 to 100 times the reported effective levels of both vancomycin and tobramycin. At minimum 2-year follow-up, the investigators reported no evidence of active infection in any patient.
Antibiotic-loaded autologous cancellous bone grafting also has shown positive results. Chan and colleagues in 1998 combined iliac cancellous bone grafts with piperacillin and/or vancomycin and implanted the mixture at the site of similar infected osseous defects. All fractures in the 36 study patients went on to union within 4 to 5 months with the only complications reported as skin rashes. In a more recent study by the same group, Chan and colleagues evaluated 96 patients for infected tibial nonunions treated with local antibiotic bead therapy and staged antibiotic-loaded versus pure autogenous bone graft. In the antibiotic-loaded group, the infection-eradication rate was 95% at more than 4 years’ follow-up with a 100% union rate versus 82% eradication and 98% union rate in the pure autogenous group. A more recent study by Khoo and colleagues reported no early infections after the insertion of antibiotic-loaded iontophoresed segmental allografts for various orthopedic limb salvage surgeries with a mean follow-up of 51 months. More studies are needed to evaluate the local antibiotic concentration level and the effect this has on eventual bone graft incorporation and bone healing. In one study comparing bone healing with autogenous cancellous bone grafting with and without admixed tobramycin, there was no effect in bone healing with large concentrations of local tobramycin. Given the differences in bone and antibiotics used in addition to loading method and dosing, firm conclusions cannot be drawn from the literature at this time and more clinical comparative studies are needed to guide future use.
Bone Graft Substitutes or Extenders
Bone graft substitutes, such as calcium sulfate, calcium phosphate, hydroxyapatite, and tricalcium phosphate, have gained interest because they are osteoconductive and are compatible with and can promote the regeneration of bone during the time of material degradation. They are also desirable because they avoid the risk of transmitting disease pathogens associated with the use of allograft. Many of these products have Food and Drug Administration–approved treatment indications for bone void filler and are readily available in most operating rooms. However, they all show a rapid release of the antibiotic at a relatively uncontrolled rate.
Of the bone graft substitutes, calcium sulfate is used most commonly in the clinical setting as an antibiotic delivery vehicle. It is commercially available and was first used as bone defect filler in 1982. Its advantages as bone void filler include its steady and gradual resorption, osteoconductive properties, lack of immunogenic side effects, and consistent clinical record. Blaha further demonstrated that calcium sulfate supports the infiltration of new blood vessels and osteogenic cells, and prevents ingrowth of soft tissue. Water-soluble antibiotics can be incorporated into the crystalline structure rather easily, although as mentioned previously, the antibiotic in this form is often eluted at an uncontrolled rate. Tobramycin, which is generally effective against the most common species responsible for osteomyelitis, can be incorporated to produce an extremely high local concentration as the pellets are resorbed. The most appropriate antibiotic dosage regimen remains unclear; however, common formulations used clinically have ranged from approximately 1 g vancomycin and/or 1.2 g tobramycin or gentamicin per 25 g calcium sulfate. Even in infections when certain organisms are not sensitive to tobramycin at levels obtainable with systemic therapy, the high local concentrations released by the calcium sulfate pellets may be effective at eradicating the infection.
Several human and animal studies have demonstrated the safety and efficacy of antibiotic-loaded calcium sulfate beads with infection-eradication rates of more than 90% ( Fig. 5 ). In a recent prospective randomized clinical trial, McKee and colleagues compared tobramycin-loaded calcium sulfate pellets with PMMA and demonstrated similar infection-eradication rates and new bone growth with the clinical benefit of requiring fewer subsequent procedures. A recent in vitro study reported that calcium sulfate as an antibiotic carrier is at least equivalent, and potentially superior to PMMA, in inhibiting bacterial growth in liquid and agar cultures. There is building evidence that antibiotic-loaded calcium sulfate pellets are a safe and effective alternative to PMMA beads, obviating the added morbidity of a subsequent procedure for removal of beads.
Calcium sulfate also has been used with other materials, such as calcium hydroxyapatite (HA), in composite carriers. The primary advantage of HA over other carriers is that it is slowly replaced by new-forming bone, which might reduce the requirement for additional reconstruction. One study used equal quantities of calcium sulfate and HA in a composite delivery system and found, as compared with calcium sulfate alone, the addition of HA slowed down the resorption and maintained serum levels of antibiotic at 4 weeks better than calcium sulfate alone. Korkusuz and colleagues also found that an HA-ceramic composite had consistently higher serum levels of gentamicin than antibiotic-loaded PMMA and resulted in infection eradication at 7 weeks in a rat osteomyelitis model. It is hypothesized that the porous structure of HA allows the infiltration of calcium sulfate, allowing a more sustained release as compared with the carriers in isolation.
The disadvantages of calcium sulfate include potential cytotoxicity, high rate of resorption, and rapid elution rates, all of which have been studied in vitro. Furthermore, calcium sulfate has a low mechanical strength, rendering it unsuitable where load bearing is required. Calcium sulfate is known to generate an inflammatory and osmotic response that may result in increased fluid in the wound bed, and may result in opaque drainage from the incision that can be mistaken for purulence or recurrence of infection. This has been reported in the literature; however, this is most commonly inconsistent with infection, and more likely an osmotic effect, as bacterial cultures are typically negative and wound healing occurs with standard dressing changes. For these reasons, this material should be used carefully in the setting of inadequate soft tissue envelope.
As discussed, biodegradable carriers are theoretically advantageous for their reduced risk of secondary infection and lack of need for a secondary surgery for removal. Furthermore, in instances in which there is little potential space to manage following irrigation and debridement, space-occupying carriers such as beads are not desirable and can make closure more difficult. This clinical scenario is typical in osteomyelitis, where the maintenance of a soft tissue envelope and long-term dead-space management is not required. The most widely used and studied biodegradable natural polymer is the collagen sponge. It is most commonly composed of a solid mesh of collagen-based material produced from sterile animal skin or the Achilles tendon. In Europe, for example, it has been commercially available for more than a decade and is produced from sterilized bovine tendon in which gentamicin is suspended. Collagen is a major component of the native connective tissue and is a structural component of virtually all organs, which makes the implant both biocompatible and nontoxic. Furthermore, its variable drug-release profile and ability to attract and stimulate the proliferation of osteoblasts and promote mineralization make it an ideal delivery vehicle. Collagen fleece kinetic studies have demonstrated a three-phase release of antibiotics. This elution profile is largely due to the distinct structure of the collagen and the porosity of the fleece. Rapid release occurs soon after implantation due to the porous structure of the fleece with an intermediate release phase following due to the partially closed pores to which the antibiotic is enclosed, and finally a prolonged release phase due to the encasement of the drug within the fibrillar collagen structure. Kinetic studies have quantified that 95% of gentamicin can be released in the first 1.5 hours compared with 8% from PMMA beads over the same time period. Elution can be elongated by slowing the degradation rate of the fleece or sponge by macrophages. This can be accomplished by using hydrophilic gentamicin sulfate in combination with hydrophobic gentamicin crobefate resulting in prolonged release over 10 days.
As compared with other biodegradables, antibiotic-loaded collagen delivery systems have been evaluated in several clinical studies. Ascherl and colleagues demonstrated a 94% infection-eradication rate in 67 patients with posttraumatic and postoperative osteomyelitis treated with a gentamicin-loaded collagen sponge. Other studies have reported infection-eradication rates of 93% to 100% with the addition of intravenous antibiotics.
Gentamicin-loaded collagen sponge shows up to 600 times the minimum inhibitory concentration (MIC) as compared with PMMA beads at 300 MIC, and has been noted to be an effective delivery vehicle for up to 28 days in a rabbit model. In a comparison study, Hettfleisch and Schottle demonstrated a superior pharmacokinetic profile with complete elution of the antibiotic in collagen fleece as compared with PMMA. In the broader literature, a recent meta-analysis was conducted investigating nearly 7000 patients who had gentamicin-loaded collagen sponges placed for the prophylaxis of surgical site infections in several different surgical settings. The study concluded that the implants decreased the rate of surgical site infections. Further investigation and characterization of the efficacy and safety of collagen sponges in orthopedic surgery are under way so they can be recommended and approved as a delivery vehicle in the United States.
More recent developmental work has focused on bioabsorbable gels as antibiotic delivery vehicles. Advantages of a gel include release of 100% of the carried antibiotics as it is degraded, and improved and more immediate distribution of antibiotic throughout the wound as opposed to discrete pockets around beads.
A recent in vitro study tested a sterile phospholipid gel containing 1.88% vancomycin and 1.68% gentamicin by weight. The study demonstrated local delivery of antibiotic by a bioabsorbable gel achieved complete wound coverage and was more effective in reducing bacteria than the commonly used antibiotic-PMMA bead depot in a contaminated defect study in a rat model. The results of the study showed high initial concentrations that slowly declined over several days, and, in a supplemental unpublished study using a rabbit model, maintained persistent local tissue levels of antibiotics greater than the MIC for more than 14 days. The proposed strength of the gel is the ability of local antibiotic levels to rapidly exceed the MIC and minimum biofilm, eradicating concentrations of bacteria within the wound followed by a sustained release to combat and eradicate quiescent cells that are not as susceptible to the antibiotics at subtherapeutic levels. The gel also provides immediate drug delivery to the entire wound contact area, with no need for the active drug to elute and diffuse through the wound bed. Bioabsorbable gel shows promise as a capable delivery vehicle and translational research is currently under way to further evaluate its safety and efficacy.
Administrations in Aqueous Solution
Antibiotics can be effectively administered in aqueous solution as well. Although antibiotic (as well as antiseptic) solutions have been used for many years, there are clinical data that suggest this method of delivery is effective. A significant positive impact was shown in a series of patients with shoulder arthroplasty, where the rate of infection was decreased by 1 order of magnitude with intra-articular injection of aqueous gentamicin performed after wound closure. This method of administration also has been effective in a rat model, where aqueous gentamicin was significantly better for infection prophylaxis after placement of a metal implant and contamination with S aureus , and this effect was further improved by concomitant administration of systemic antibiotics. This method of delivery is distinct from antibiotic solution for irrigation, as the aqueous antibiotic is injected into the surgery site after the wound is closed. An 18-gauge needle is used adjacent to the incision to inject the solution throughout the wound. This technique allows the solution to ideally infiltrate the interstices or depths of the wound. In practice, the solution has been made in a concentration of 2 mg/mL (80 mg aminoglycoside in 40 mL injectable saline). Recent data, which suggested lower local cytotoxicity with tobramycin compared with gentamicin, has prompted the switch to tobramycin in this application.
The advantages of delivering antibiotic in aqueous form are several, including cost, as there is no need for a specialized vehicle of delivery. The antibiotic drugs are approved and ready for use. Also, the wound encounters a higher maximum antibiotic concentration, potentially improving antibiotic efficacy.
Drawbacks to this method may include poor sustained antibiotic level; however, other vehicles for delivery, such as collagen sponge, have been shown to elute the drug very quickly, then potentially become a foreign body at the surgical site. Thus, this method of antibiotic delivery may be more effective in a situation in which sustained delivery is not required, such as prophylaxis of surgical site infections.
Local Antibiotic Pharmacokinetics and Safety
As stated previously, local antibiotics have the advantage of high local concentrations (thus efficacy at the surgical site), and low systemic concentrations (less risk of systemic side effects). Each of the carriers described previously elutes antibiotic at a different rate, resulting in variable potential risk for local cytotoxicity. In an in vitro model, bone graft and demineralized bone matrix (DBM) were shown to elute 70% and 45% of their antibiotic load by 24 hours, and negligible amounts were detected at 1 week. In this study, PMMA cement was shown to elute 7% at 24 hours, with continued elution detected at 14 days. Calcium sulfate was shown to release 17% of its antibiotic by 24 hours, with trace amounts detected at 3 weeks. In other calcium sulfate elution studies, approximately 58% of the contained antibiotic is released during the first 24 hours in in vitro elution studies, and 20% released within the first 24 hours in an in vivo rabbit model. Thus, bone graft and DBM may be best if only short duration of antibiotic coverage is needed, where calcium sulfate or bone cement may provide more prolonged coverage. Aqueous antibiotic solution would be expected to provide the shortest duration of coverage in the local tissue, as there is no vehicle of delivery.
Local levels as measured in wound exudate have been shown to be up to 300 μg/mL with gentamicin in PMMA. Anagnostakos and colleagues demonstrated that peak antibiotic levels for gentamicin and vancomycin from PMMA beads and spacers as measured in the drainage is highest on day 1, which then decreases significantly; however, antibiotic levels did persist through the 13 days tested. They concluded that beads eluted antibiotics more quickly and with higher peak concentration than spacers due to greater surface area, but that there was considerable variability between subjects for a given method of delivery, noting a lack of consistency in the elution profile.
Regarding local toxicity and safety, Rathbone and colleagues demonstrated levels of toxicity for many commonly used antibiotics. These data are helpful for determining appropriate concentrations that are acceptable in the wound cavity, regardless of the delivery method. On evaluating osteogenic cell viability and activity as measured by alkaline phosphatase activity, wide variability was noted, even within families of antibiotics ( Fig. 6 ). Rifampin, tetracyclines, and ciprofloxacin were noted to be particularly cytotoxic, with considerable decrease in cell number and activity at concentrations of even 100 μg/mL. Tobramycin, vancomycin, and amikacin were noted to be least cytotoxic. Of note, cell number and activity were measured at time points of 10 and 14 days for various concentrations of each antibiotic. Although the response of osteogenic cells to these concentrations of antibiotics is useful information, it should be noted that these are measured in response to sustained levels of antibiotics, and in the clinical setting, the antibiotic levels are highest at early time points and then often decrease rapidly.
Regarding the systemic levels achieved after local administration, Livio and colleagues showed calcium sulfate with 4% tobramycin (262 or 524 mg tobramycin per 10 or 20 g calcium sulfate, respectively) resulted in systemic levels of tobramycin that fell well below the accepted safe serum level (2 mg/L) within 24 hours. However, based on their projected data, sustained potentially toxic concentrations were predicted by simulations with these doses at renal failure stages 4 and 5. Thus, even standard doses of tobramycin-loaded calcium sulfate should be used with caution in the setting of advanced renal failure.
Current treatment algorithms of prolonged systemic intravenous antibiotic treatment are becoming increasingly less effective with increasing bacterial resistance in addition to increasing patient costs and morbidity. A multidisciplinary approach with the application of various techniques will be required to treat orthopedic-related infections. At the forefront of this initiative are biofilm-disrupting treatments that break up biofilm layers into a single-celled planktonic form that are more susceptible to antibiotic treatment. Laser-generated shockwaves have been tested in vitro that use mechanical energy have been found to remove 97.9% of P aeruginosa biofilms on nitinol stents. Furthermore, electrical stimulation on orthopedic implants is being studied to enhance detachment of the biofilm. Studies have shown that electrical current is able to enhance the detachment of biofilms growing on stainless steel implants infected with S aureus and Staphylococcus epidermis . One study reported that the MIC of S epidermidis to gentamicin was decreased by 50% when infected stainless steel pegs were pulsed with electromagnetic energy. Thus, applying an electric current in combination with antibiotic therapy may help patients avoid the morbidity and cost of going through major revision surgeries and chronic suppressive antibiotic regimens.
The use of rifampin has been evaluated, including both local and systemic routes of administration. It has been demonstrated effective at penetrating bacterial biofilm in both in vitro and animal models. One problem is that rifampin is relatively cytotoxic, resulting in a decreased osteoblast cell number and alkaline phosphatase activity at a concentration of less than 100 μg/mL. Further research is required to determine local versus systemic efficacy, use with other antibiotics, and safe local concentrations.
Use of nonantibiotic agents, such as d -amino acids, may result in degradation of biofilms and make them more susceptible to antibiotic activity. Bacillus subtilis has been found to produce this combination of d -amino acids as a signal to trigger break down of biofilms. However, this d -amino acid combination is not specific to Bacillus . In an in vitro setting, Losick demonstrated increased biofilm breakdown, and Wenke and colleagues showed that antibiotics were more effective against S aureus and P aeruginosa biofilm when administered with d -amino acids. Although only investigated in vitro at this point, the development of d -amino acids may be an effective adjunct in the treatment of biofilms and established infections.
The interest and research being devoted to the development and design of the ideal local antibiotic delivery continues to increase with the ultimate goal of obviating the need for systemic antibiotics for the treatment of many musculoskeletal infections. These local antibiotic therapies theoretically spare the morbidity and mortality associated with system drug toxicity and may be more effective at their targeted site than systemic therapies.